Organizational de-structuring? Latour’s potential contribution to the critical realist – pragmatist dispute

ABSTRACT

This article explores a key difference that Elder-Vass has identified between critical realism and pragmatism: their divergent views on the viability of the concept of social structure. Noting that this is also a point of dispute between critical realists and Actor-Network Theorists, I try to contribute to the debate about social structure by focusing on the question of whether organizations are structures, drawing on the views of a pragmatist-leaning ANT - Bruno Latour - to evaluate those of a critical realist - Dave Elder-Vass. I suggest that Latour's arguments can be used to identify two challenges for Elder-Vass's approach, questioning both whether an organization has a singular structural form and whether organizations have determinable powers and impacts. I then consider whether Latour's approach is vulnerable to critical realist arguments that perspectives which deny the existence of social structure cannot account for long-term stabilities, fail to connect different cases, and are individualistic in character.

KEYWORDS:

• Critical realism
• Elder-Vass
• Latour
• organization
• pragmatism
• structure

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 I really appreciate feedback from Christoforos Bouzanis and Yufan Sun on a version of this paper, as well as discussions with Sharani Osborn and Angélica Thumala Olave.

2 Latour is pretty clearly not a ‘naturalist’ in any typical sense of the word, and his thought might be considered a-naturalistic insofar as he does not wish to give any special priority to scientific inquiry in determining the contents of the world. This failure to prioritize science clashes fairly straightforwardly with Dewey’s later naturalism (see Godfrey-Smith 2002 for discussion) but may have some degree of affinity with James’ views (for one discussion of the limits of James’ naturalism see Gale 1997).

3 An important issue here is whether or not social structures should be understood as partially constituted by agents’ conceptions and actions.

4 Roughly speaking, for Latour an association is a connection made by an actor with other actors (human or non-human) which is invoked as part of the initial actors’ action.

5 Elder-Vass refers to roles as social positions (2010, 10), so this seems like a reasonable terminological move.

6 One might see a parallel with Latour’s Tardeian argument that what may be wrongly seen as a singular general property of the collective is actually the product of a blurring together of a range of different people’s ways of thinking and acting (Latour 2010, 149–50).

7 This suggests that there are intrinsic complications to role expectations which are not captured in statements such as the following: ‘Role descriptions implicitly or explicitly specify norms that define how an incumbent of the position concerned must relate to other members of the organisation and also how they must relate to outsiders when acting on behalf of the organisation.’ (Elder-Vass 2010, 153). The emphasis on how incumbents ‘must’ relate perhaps sounds like an exasperated manager trying to insist on their version of roles and norms.

8 These actors could, in Elder-Vass’s terms, be reasonably described as representing the organization. Nevertheless, the ‘organization’ itself does not tell them what to do, so there is then a question of their decision-making and the information on which this is based.

9 Further, for Latour the very question of what counts as a successful or effective action is up for grabs in the activity of organizing, rather than a clear-cut benchmark that outcomes can be judged against (2013, 395).

10 Baker and Modell (2019) make an admirable attempt to link theory and research by using critical realist concepts such as ‘norm circle’ to analyse the extent of the performativity of ideas of corporate social responsibility in relation to the activities of an Australian packaging company over time. They contend that there are (at least) three norm circles influencing the behaviour of the company and its employees: one relating to normative management knowledge, a second relating to customer demands and a third relating to labour rights. Their analysis is thus attempting to show that there are norm circles operating within organizations even if the norm circles they consider are not directly linked to role positions. To offer an unsatisfactorily brief response to this I would note that: (i) the norm circles they consider seem to be broader than the institution, rather than based in a tightly-knit set of actors inside the company; (ii) one of the norm circles, customer demands, isn’t characterized as involving mutual and consistent sanctioning of all the relevant actors by one another, but as involving a set of actors whose demands converge with one another in their dealings with one actor - the packaging company - rather than their dealings with each other; (iii) another of the norm circles, labour rights, is characterized in a very broad way, with a general presentation of its purported normative commitments and an acknowledgement of the likely differences of views between workers that suggests that this is not the kind of tightly-connected, mutually reinforcing set of actors Elder-Vass identifies as characteristic of norm circles.

11 At the other extreme, the presence of any meaningful contingency at all could be denied. I adopted a position in this territory in Kemp and Holmwood (2012) but, as the nature of the current article implies, I am uneasy about the defensibility of such a view.

12 A fairly recent variation on this process is the extent to which, during Donald Trump’s US presidency, aides sometimes ignored his orders or ‘slow-walked’ them (Plott 2019).

13 Andrew Abbott’s pragmatist-leaning work is based in a processual ontology and has some intriguing lines of arguments about how to account for stabilities within such an approach (see e.g. Abbott 1995). Thanks to Yufan Sun for pointing out the links between Latour and Abbott.

14 There are aspects of Latour’s work, such as his discussion of forms of standardization, which arguably push in a different direction (see e.g. Latour 2005). But these are hard to square with the more contingentist aspects of his thought.

15 I use the term ‘non-structural processualism’ because critical realist approaches could reasonably be said to analyse ongoing processes, but consider social structures to be a component of these, which Latour (obviously) denies.

16 I hope to explore these issues as part of a book-length treatment of Latour’s critique of sociology.

Additional information

Notes on contributors

Stephen Kemp

Stephen Kemp is a senior lecturer in sociology at the University of Edinburgh. He writes about a range of social theory issues in the areas of epistemology and ontology.